In kindergartens and nurseries, a musical classroom is an especially important activity for improving children’s self-expression abilities. This paper proposes a children’s music education support system that uses sensor data, video images, and sound data. Our system can give suggestions to a teacher by observing and analyzing a child’s actions. The proposed system focuses on the motion of a baton equipped with an LED and an accelerometer. A child shakes a baton accompanied by a piano. Marker track- ing results and accelerometer output are represented graphically: the horizontal and vertical axes are time and observed values, respectively. An experiment at a kinder- garten showed that inexperienced teachers recognize more detailed and rich informa- tion from data in a graphical format than observation with the naked eye.